[erlang-questions] RE : Erlang now in the railway industry

BETIS Alexandre <>
Mon Sep 11 17:19:11 CEST 2006


Hi,

> 
> How did you convinced your hierarchy and your customer ?
>

This question pops up quite often. It is pivotal to 
the success of Erlang in the future, hence I will 
take some time to answer it.

The short, ugly truth is that I didn't. I half-lied. 
I pretended Erlang was just a C library dedicated to the 
writing of high availability systems. In fact, this is 
partly true. As Mr Spock would say, "Why lie when you can 
just omit?".

It was a calculated risk. It worked and proved profitable
for everyone.

Now, for the long story,

Convincing your hierarchy
-------------------------

I did not convince my hierarchy. My results did.
Thanks to Erlang, my estimates were accurate. 
When I said this feature worked, it actually
did! Consequently, virtually all change requests 
where specification-level stuff. Interestingly, this 
is something project managemenent rather liked. Such
claims can be traded against other stuffs.

Convincing your customer
------------------------

I did not convince my customer. The software did.
A non-crashing software is something that turns
your customer into your best friend. The trick
is to deliver it quickly, and track requirements
carefully. Agile techniques coupled with the Erlang 
architecture are a powerfull combination to achieve 
this (but this exceeds the scope of this answer).

Convincing yourself
-------------------

You did not ask, but the most important thing is 
to convince yourself that the risk is worth being 
taken. This is the hardest part because you're
going to be jumping freefall with a new untested
gear.

There are three reasons that had me jump with
the Erlang (c) parachute:

1/ I have a Comp Sci background. I was familiar 
with functional programming Languages and knew
they were much much better for the programmer's
mental health.

2/ The test_server. Erlang was the first platform
that came with a regression test suite. After
running it on the target platform repeatidly for 
a few days, I was much more confident in this
technology (granted, it crashed after three days.
The disk was full with test reports. Well, doh...)

3/ For various reasons, including past projects
gone awry because of overhyped technology,
I was /annoyed/ (I don't want to use the p word
in a well kept forum). Wanting to kick butt
can be a great motivation.

Bottom line
-----------

You cannot convince with just technology and talks.
There is too much of marketing firepower in front 
of us to win a battle of words.

Only demonstrated profitability will do that. Pick 
up one of your assigned project, not too big (e.g. a 
module), make it happen, and expand from there. 

It may take a while to find an opening window, but if 
the technology really fits the needs of the organization,
it will pop up sooner or later. Grab it. But like all
pioneers, you will be alone.

Cheers,

-- Alex




" Ce message et ses pièces jointes sont confidentiels et exclusivement réservés à leurs destinataires qui s'assureront de l'absence de virus. CSEE Transport n'assume aucune responsabilité au titre de ce message s'il est altéré, déformé, falsifié, indûment utilisé par des tiers ou encore s'il a causé des dommages ou pertes de toute nature.

This e-mail and any attached files are confidential and exclusively intended for its addressees who will conduct appropriate virus checks. CSEE Transport shall not be liable for the message if altered, changed, falsified, unduly used by third parties nor for any damage or loss."





More information about the erlang-questions mailing list